This application is submitted in response to PA-06-101, "Research on Clinical Decision Making in Life- threating Illness (R21)". Most Americans are unable to participate in care decisions at the end of life due to life-threatening medical conditions. In fact, proxy decision makers, usually the closest family member, usually make the end of life treatment decsions with healthcare provdiers. Unfortunately, the research literature on proxy decision making reveals poor agreement between elderly patients and their designated proxy decision makers regarding hypothetical poor-health scenarios. The reasons for the discordances between proxies and patients are not well understood. No intervention designed to improve the accuracy of proxy decision makers has been effective. This high discordance poses a serious challenge for the bioethical concern with patient autonomy, the ideal of substituted judgment and the strength of the shared decision making model. The proposed research will study the treatment preferences of elders and their proxies and factors that influence their treatment preferences. Two hundred pairs of elders and their designated family proxy decision makers (elder-proxy dyads) will be recruited through area continuing care retirement communities (assisted living and independent housing), senior centers, and display ads in publications that target elderly populations. Telephone interviews, about 45 minutes long, will be conducted with elders and their proxies (separately), using questionnaires that assess preferences for life-prolonging treatment, end of life values, social support factors, religious/spiritual orientation, health, and demographic characteristics. Data analysis, primarily multiple regressions, will assess the treatment preferences of elders and their proxy decision makers. Factors influencing the treatment preferences of elders and their proxies will be identified. The magnitude of elder-proxy discordance will also be determined as well as factors associated with this discordance. Identification of the factors contributing to the discordance in treatment preferences between elders and their proxies will have important clinical, educational and research applications. A better understanding of these factors will ultimately improve the shared decision making done between healthcare providers and families at the end of life for decisionally-incapacitated patients. [unreadable] [unreadable] [unreadable]